Báo cáo khoa học: Coexpression, purification and characterization of the E and S subunits of coenzyme B12 and B6 dependent Clostridium sticklandii D-ornithine aminomutase in Escherichia coli potx

Coexpression, purification and characterization of the E and S subunits
of coenzyme B
12
and B
6
dependent
Clostridium sticklandii
D
-ornithine
aminomutase in
Escherichia coli
Hao-Ping Chen
1
, Fang-Ciao Hsui
1
, Li-Ying Lin
1
, Chien-Tai Ren
2
and Shih-Hsiung Wu
2
1
Institute of Biotechnology and Department of Chemical Engineering, National Taipei University of Technology, Taipei, Taiwan;
2
Institute of Biological Chemistry, Academia Sinica, Nankang, Taipei, Taiwan
D
-Ornithine aminomutase from Clostridium sticklandii
comprises two strongly associating subunits, OraS and
OraE, with molecular masses of 12 800 and 82 900 Da.
Previous studies have shown that in Escherichia coli the
recombinant OraS protein is synthesized in the s oluble form

and OraE a s inclusion bodies. Refolding experiments also
indicate that the interactions between OraS and OraE and
the binding of either pyridoxal phosphate (PLP) or aden-
osylcobalamin (AdoCbl) p lay i mportant roles in the
refolding process. In this study, the DNA fragment con-
taining both genes was cloned into the same expression
vector an d coexpression of the oraE and oraS genes was
carried out in E. coli . The solubility of the coexpressed
OraS and O raE increases with decreasing isopropyl thio-
b-
D
-galactoside induction temperature. Among substrate
analogues tested, only 2,4-diamino-n-butyric acid displays
competitive inhibition of the enzyme with a K
i
of
96 ± 14 l
M
. Lys629 is responsible for the binding of PLP.
The apparent K
d
for coenzyme B
6
binding to
D
-ornithine
aminomutase is 224 ± 41 n
M
as measured by equilibrium
dialysis. T he m utant protein, OraSE–K629M, i s s uccessfully

expressed. It is catalytically inactive and unable to bind PLP.
Because no coenzyme is involved i n protein folding during
in vivo translation of OraSE–K629M in E. coli, in vitro re-
folding of the enzyme employs a different folding mechan-
ism. In both c ases, the association of the S a nd E subunit is
important for
D
-ornithine aminomutase to maintain an
active conformation.
Keywords: adenosylcobalamin; B
12
;
D
-ornithine amino-
mutase.
D
-Ornithine aminomutase from Clostridium sticklandii
catalyzes the r eversible interconversion of
D
-ornithine into
2,4-diaminopentanoic acid [1]. It comprises two strongly
associating sub units, O raS a nd OraE , w ith m olecular
masses of 1 2 800 and 82 900 Da. Two diff erent coenzymes,
pyridoxal phosphate (PLP) and adenosylcobalamin ( Ado-
Cbl), are involved in this enzymatic reaction. The genes
encoding
D
-ornithine aminomutase, oraE and oraS,have
been cloned, sequenced, and expressed in Escherichia coli
[2]. The recombinant OraS protein was synthesized in a

soluble homogeneous form, but the majority of OraE
protein was produced in the form of inclusion bodies. T he
enzymatic a ctivity could be restored after a refolding step.
However, OraE could not be properly folded in the absence
of OraS and coenzyme. These observations indicate that the
binding of AdoCbl or PLP and the interactions between
OraS an d OraE play important roles i n t he OraE refolding
process. The strong interaction between the E and S
subunits of the enzyme was first reported by Barker &
Stadtman [3]; Barker discovered glutamate mutase, which is
also composed of weakly interacting E and S components.
The correlation between these interactions and protein
folding is not clear.
As protein refolding is labor intensive and time consu-
ming, an efficient expression system to produce large
amounts of soluble proteins in a short time is required.
Instead of expressing the oraE and oraS genes separately,
the DNA fragment containing both genes was cloned i nto
the same expression vector under the control of the T7
promoter, and coexpression of oraE and oraS genes was
carried out in E. co li. Meanwhile, the extent of the
involvement of AdoCbl or PLP in the in vivo folding
process w as also investigated. We now describe the
construction, coexpression, and purification of the apo-
enzyme of
D
-ornithine aminomutase, together with the
temperature effect o n p rotein expression and characteriza-
tion of the recombinant proteins.
Materials and methods

Materials
AdoCbl was obtained from Sigma. Micro Dialysis tube,
Q-Sepharose High Performance anion-exchange medium
and Phenyl-Sepharose High Performance hydrophobic
interaction g el medium were from Amersham Biosciences.
Restriction endonucleases, BamHI, SpeI, an d NcoI,
Correspondence to H P. Chen, Institute of Biotechnology and
Department of Chemical Engineering, National Taipei University of
Technology, 1, Sec 3, Chung-Hsiao East Road, Taipei 106, Taiwan.
Fax: +886 2 27317117, Tel.: +886 2 27712171 ext. 2528,
E-mail:
Abbreviations: AdoCbl, adenosylcobalamin; IPTG, isopropyl thio-b-
D
-galactoside; PLP, pyridoxal phosphate.
(Received 3 0 June 200 4, revised 26 August 2004 ,
accepted 17 September 2004)
Eur. J. Biochem. 271, 4293–4297 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04369.x
DNA-modifying enzymes, and Ex Taq DNA polymerase
were purchased from TaKaRa (Otsu, Japan). The E. coli
strain BL21(DE3) codon plus was from Stratagene.
1,4-Diaminobutane and (R,S)-2,4-diamino-n-butyric acid
were from Sigma. 4-Am inopentanoic acid, and 2,5-diamino-
pentanol were the k ind gift from T L. Shih ( Department of
Chemistry, Tamkang University, Taiwan). All c hemicals
used were of molecular biology grade o r higher.
Construction of expression vector poraSE
A pair of oligonucleotides, 75 and 44 (Table 1), was
designed using the nucleotide sequence of the ora genes in
order to f acilitate t he amplication by PCR. An NcoIsitewas
introduced into the start of the oraS gene and a BamHI site

into t he e nd o f the oraE gene. Genomic DN A w as purified
from C. sticklandii by phenol/chloroform extraction m eth-
ods [4]. The coding regions for the S and E subunits of
D
-ornithine aminomutase were then amplified by PCR
using c lostridial genomic DNA as template. Amplification
was achieved using 30 cycles at the f ollowing temperatures:
95 °C for 30 sec, 50 °C for 1 min, and 72 °Cfor4min.
Finally, the reaction was maintained at 72 °Cfor5min.
The PCR products were gel-purified, restricted with NcoI
and BamHI, and ligated with NcoI/BamHI restricted pET-
28a vector. The ligation mixture was used to transform
E. coli DH5a. The plasmid that carried the oraS and oraE
genes in the correct orientation was designated poraSE.
Isopropyl thio-b-
D
-galactoside induction temperature
and small-scale expression
To facilitate the over-expression of the ora genes, poraSE
wasusedtotransformE. coli BL21(DE3) codon plus.
Cultures were first grown at various temperatures in
500 m L L uria–Bertani medium containing kanamycin
(30 mgÆL
)1
). After isopropyl thio-b-
D
-galactoside (IPTG)
induction and expression, the cells were harvested by
centrifugation and resuspend ed in 1 5 mL of 50 m
M

phos-
phate buffer, pH 7.0. The cells were ruptured by sonication
and cell debris was removed by centrifugation at 25 0 00 g
for 1 5 min at 4 °C. To examine the solubility of th e over-
expressed proteins, 20 lL o f supernatant was taken for
analysis by SDS/PAGE. The insoluble f raction and cell
debris from 1 mL overnight culture were collected by
microcentrifugation at 13 000 rpm f or 5 min and dissolved
in 100 lL SDS/PAGE loading buffer; 10 lLwastakento
analyse by SDS/PAGE.
Large-scale protein expression and purification
Cultures were grown at 25 °C by inoculating a 5 mL
overnight c ulture into 4 L of Luria–Bertani m edium
containing kanamycin (30 mgÆL
)1
). Incub ation was contin-
ued until the culture reached an attenuance of 0.6–0.8 at
600 n m, at which point the temperature was lowered to
20 °C and expression was induced by the addition of
200 mgÆL
)1
IPTG. After overnight incubation, cells were
harvested by centrifugation at 4000 g for 10 min. The cells
were then stored at )20 °C.
All purifi cation steps were performed on ice or at 4 °C.
In a typical purification, 15 g of cells (wet weight) were
resuspended in 3 0 mL of 5 0 m
M
Tris/Cl buffer, pH 9.0. The
cells were ruptured in a v olume o f 3 0 mL by s onication. Cell

debris was removed by centrifugation at 25 000 g for
15 min. The supernatant was applied directly to a
2.6 · 20 cm Q-Sepharose Fast Flow anion-exchange col-
umn equilibrated with 10 m
M
Tris/Cl buffer, pH 9.0.
Protein was eluted with a 600 mL gradient from 0 to
0.5
M
KCl. The flow rate was 1 mLÆmin
)1
; 5 mL fractions
were collected. A ctive fractions were pooled and brou ght to
25% saturation i n ammonium sulfate by slow addition of
solid. The precipitate was removed by centrifugation at
25 000 g for 30 min and the supernatant was applied
directly to a Phenyl-Sepharose High Performance hydro-
phobic interaction column (2.6 · 25 cm) equilibrated w ith
10 m
M
Tris/Cl buffer, pH 9.0, co ntaining 1
M
(NH
4
)
2
SO
4
.
After w ashing the column with 100 mL of the same buffer,

the e nzyme was eluted with a linear, descending gradient of
ammonium sulfate in 1000 mL o f buffer. The flow rate was
1mLÆmin
)1
; 10 mL fractions were collected. Active frac-
tions were pooled and concentrated to 15 mL by ultrafil-
tration in a stirred cell fitted with a YM-3 membrane. The
protein solution was stored at )80 °C in the presence of
50% (v/v) glycerol.
Mutant construction
The construction of mutant poraSE-K629M was carried
out using r ecombinant P CR [5]. Two overlapping, comple-
mentary oligonucleotides were designed to introduce the
mutagenic sequence. A 1.8 kb and 700 base pair region of
the oraE gene were PCR amplified using poraEX as
template and oligonucleotide pairs 40/66 and 41/67 as
primers. Both PCR products were gel-purified and assem-
bled in a second round of PCR using oligonucleotides 40
and 41 as primers and cotemplates. The PCR product was
purified, restricted with SpeIandBamHI, and ligated with
SpeI/BamHI-restricted poraSE vector. The resulting p las-
mid was designated poraSE-K629M. The DNA fragment
amplified by PCR was resequenced by automated methods
(Mission Biotech Co. L td., Nankang, T aipei, Taiwan; A BI
3730 XL DNA A nalyzer, Applied Biosystems, CA, USA)
to confirm that no unwanted mutation h ad been introduced.
The procedures for production and purification of the
mutant protein were the same as those of the wild-type.
Protein determination and enzyme assay
Protein concentrations were determined by the method of

Bradford using bovine serum albumin as standard [6]. The
Table 1. PCR primer names and seq uences.
Primer
name Sequence
21 GGGTCTAGAATGGAAAAAGATCTACAGTTAAGA
33 CCGGAATTCTTATTTCCCTTCTCTCATCTC
40 GCGCGCCATGGAAAAAGATCTACAGTTAAGA
41 GGGGGATCCCCATAATCCACTCCACCTGCTAAA
44 GGGGGGGATCCT CATTATTTCCCTTCT
66 AATACCGCCATGTATAATATCTATTACTTC
67 GTAATAGATATTATACATGGCGGTATTGAA
75 GGGGGGGCCATGGAAAGAGCAGACGATTT
4294 H P. Chen et al.(Eur. J. Biochem. 271) Ó FEBS 2004
assay k it was obtained f rom B io-Rad, H ercules, CA, USA.
A rapid spectrophotometric method was used to assay
D
-ornithine aminomutase activity [7]. The assay couples
the reduction of NADP
+
to form 2,4-diaminopentanoic
acid through the action of NAD
+
/NADP
+
-dependent
2,4-diaminopentanoic acid dehydrogenase. The K
i
value
of the competitive inhibitor, 2,4-diamino-n-butyric acid,
was determined by measuring the apparent K

m
value of
D
-ornithine at 25 , 50, 100, 200 and 400 l
M
of the inhibitor.
For the measurement of the activity of the substrate
analogues, an HPLC and NMR-based method was devel-
oped. A 1.0 mL solution in a septum-sealed vial containing
6 l
MD
-ornithine aminomutase, 0.4 m
M
PLP, an d 0.14 m
M
AdoCbl in 100 m
M
potassium phosphate buffer, pH 7.8,
was made anaerobic by purging with ar gon. A c oncentrated
anaerobic 0 .1 mL solution of substrate analog (0.5
M
)inthe
same buffer was introduced into the vial by syringe to
initiate the reaction. After overnight incubation at room
temperature in the dark, the reactions were stopped by
freeze-drying. The reaction products were separated by
HPLC on a C
18
reverse phase column with a linear g radient
of acetonitrile containing 0.1% (v/v) trifluoroacetic acid.

The substrate analogues, presumed products, and phos-
phate ion were eluted at the beginning of the run and
collected by hand. The mixture was dried by evaporation
under vacuum and redissolved in 0.4 mL D
2
Othreetimes.
The solution was transferred to an NMR tube and the
spectra were recorded at 400 MHz.
Measurement of the binding of PLP to proteins
The binding of coenzyme B
6
to
D
-ornithine aminomutase
was measured by equilibrium dialysis. About 250 lLof
30 l
M
purified proteins were loaded i nto the Micro Dialysis
tube. The protein s olutions were dialyzed against 4 00 mL of
10 m
M
Tris buffer, pH 9 .0, in the presence o f 6000, 1500,
960, 480, 300 and 150 n
M
coenzyme B
6
at 4 °Cfor4h.
Absorbance was recorded at 420 nm using an Amersham
Bioscience Ultrospec 2100 spectrophotometer; a sample of
the c orresponding dialysis buffer was used to s ubtract out

the c ontribution of unbound PLP from the absorbance of
the e nzyme. A computer program (
KALEIDA GRAPH
, Synergy
Software, Reading, PA, USA) was used to fit the data in
order to estimate the dissociation constant.
Ultraviolet–visible protein spectrum
About 16 mgÆmL
)1
proteins (wild-type or mutant OraSE-
K629M) and 3 l
M
PLP in 10 m
M
Tris/Cl buffer, pH 9.0,
were dialyzed in the dark at 4 °C, against 10 m
M
Tris/Cl
buffer, pH 9.0, containing 3 l
M
PLP for 24 h, by which
time equilibrium had been reached. Sepctra were recorded
using an Amersham Bioscience U2100 spectrophotometer;
a sample o f t he dialysis buffer was used to subtract out the
contribution of unbound P LP from the s pectra of proteins.
Results
The expression of poraSE was first carried out at 37 °Cwith
a shaking speed of 180 r.p.m. It is w orth noting t hat, alone,
OraS protein can be express ed i n a soluble form. However,
the OraE and OraS proteins were coexpressed in an

insoluble form under t he same conditions. The codon usage
difference between C. sticklandii and E. coli does not seem
to be responsible for t his result, because the E. coli strain,
Epicarian ColiÒ-Codon Plus
TM
(DE3)-RIL, contains extra
copies of the argU, ileY ,andleuW tRNA genes.
The coprecipitation of OraS and O raE might imply that
(a) the apoenzyme or O raE is not properly folded; and (b)
the noncovalent interaction between these two subunits is
strong enough to result i n the coprecipitation of OraS. In
many instances, t he folding of the desired e xpressed protein
can be improved at lower induction temperatur es [8–10]. As
shown in Fig. 1 , the solubility of the overexpressed OraS
and O raE i ncreases with decreasing IPTG ind uction
temperature. When the incubator shaking speed reduced
from 180 to 50 r .p.m., no significant difference in the
expressed protein solubility can be observed (data not
shown).
The protocol described above gave good expression of
the ora S and oraE genes. Approximately 1 5 mg of purified
protein was obtained per litre of culture. A purification
method based on chromatography on Q-Sepharose ion-
Fig. 1. The over-expres sion o f or aS and oraE at diffe rent temperatures.
(A) Supernatant fraction. (B) Prec ipitation fraction . L ane 1, marker;
lane 2, 37 °C; lane 3, 30 °C; lane 4, 25 °C; lane 5, 20 °C.
Ó FEBS 2004
D
-Ornithine aminomutase from C. sticklandii (Eur. J. Biochem. 271) 4295
exchange and P henyl-Sepharose hydrophobic interaction

matrixes was developed. In both purification steps, OraS
and OraE eluted during t he end of t he run in a well-resolved
broad peak, resulting i n protein that was nearly homogen-
eous (Fig. 2). This method of preparation proved very
reproducible, and purified enzyme could be stored in
concentrated solution in the presence of 50% glycerol for
several months, frozen at )80 °C.
A lysine residue is thought to be involved in PLP-binding
through a Schiff base linkage. Comparison of t he deduced
amino acid sequence of oraE to those of known PLP-
dependent aminomutases reveals the presence of a con-
served PLP-binding site, a lysine residue at position 629, at
the C -terminus of the OraE p rotein [11]. The binding of PLP
to
D
-ornithine aminomutase was investigated by equilib-
rium dialysis. The proteins in the Micro Dialysis tube were
equilibrated in various concentrations of PLP, and the
binding of coenzyme was measured. PLP w as bound with
an apparent K
d
of 227 ± 41 n
M
(Fig. 3).
The production and purification methods for mutant
protein, OraSE-K629M, were as described above. No
significant difference in protein solubility could be found
between wild-type and mutant protein at various IPTG
induction and expression temperatures. Perhaps not sur-
prisingly t he m utation of the L ys629 residue to Met c aused a

complete loss of catalytic activity. Meanwhile the bin ding of
PLP by mutant OraE-K629M was too weak to allow
binding constants to be determined with any accuracy, as
shown by the equilibrium dialysis experiment. The ultra-
violet–visible spectrum of wild-type and mutant enzyme is
shown in F ig. 4. The presence of an absorption maximum
at 420 nm suggests that
D
-ornithine aminomutase, as is the
case with other pyridoxal 5 ¢-phosphate dependent enzymes,
binds pyridoxal 5 ¢-phosphate via an azomethine link
between the formyl group of the c ofactor and the amino
group of a protein residue. In contrast, the absence of
absorption maximum at 420 nm of the mutant enzyme
spectrum directly demonstrates that the Lys629 residue is
responsible for the binding of PLP in
D
-ornithine amino-
mutase (Fig. 4).
High substrate specificity i s a common f eature for most
AdoCbl-dependent mutases. However, alternative s ub-
strates exit in the case of B
12
-dependent glutamate mutase
and lysine aminomutase [12,13]. The e nzymatic activity of
D
-ornithine aminomutase to four substrate analogues,
1,4-diaminobutane, 2,4-diamino-n-butyric acid, 4-amino-
pentanoic acid, and 2,5-diaminopentanol, was als o exami-
ned i n this study. Our results show that n one of them could

be catalyzed by the enzyme. Moreover, only 2,4-diamino-
n-butyric acid is able to behave as a competitive inhibitor o f
theenzymewithaK
i
of 96 ± 14 l
M
as measured by
photometric assay. The other three analogues showed
neither inhibitory potential nor suggestion of cleavage of
the cobalt–carbon bond of AdoCbl (H P. Chen, unpub-
lished results). These results suggest that the substrate
specificity of
D
-ornithine aminomutase is strict.
Discussion
The genes encoding
D
-ornithine aminomutase, oraE and
oraS, are adjacent on the clostridial chromosome. They
Fig. 2. SDS/PAGE results of samples taken after each step in the
purification of t he recombinant enzyme. Purification of OraSE (20%
gel). Lane 1, m arker; lane 2, c rude ce ll e xtract b efore I PTG i nduction;
lane 3, crude cel l extract after IPTG i nduction; lane 4, supernatant
after cell disruption by sonication; lane 5, pooled fractions after
Q-Sepharose HP chromatography; lane 6, p ooled fractions after
Phenyl-Sepharose HP hydrophobic interaction chromatography.
Fig. 3. Binding of PLP to recombinant
D
-ornithine aminomutase
measured by equilibrium dialysis.

Fig. 4. UV–visible spe ctrum of wild-type and mutant
D
-ornithine amino-
mutase. The maximal absorption at 420 nm of indic ated that pyridoxal
5¢-phosphate is bo und to the wild-type en zyme.
4296 H P. Chen et al.(Eur. J. Biochem. 271) Ó FEBS 2004
share overlapping start and stop codons which m ight lead to
transcription coupling so as to produce equal amounts of
the two proteins [2]. In the open reading frames for the oraS
and oraE genes, an E. coli ribosome-binding site is located
upstream of the initiation codon of oraS and a clostridial
ribosome-binding site on oraE. Although the different
prokaryotic Shine–Dalgarno s equences might have differ ent
affinities for ribosomes, both oraS and oraE genes are
successfully overexpressed (Fig. 5).
The strong interaction between OraS and OraE was
first reported by Barker & Stadtman [3]. OraS shows no
significant homology to other proteins in the SWISS-
PROT database. The sequence alignment results indicate
that the coenzyme-binding and catalytic domains are
located in the E subunit [2]. Unfortunately, varying the
induction temperature a nd inducer concentration had
little effect on the solubility of OraE, and any attempt to
refold OraE by itself was not successful. Although the
role of the S subunit remains obscure, it seems likely t hat
OraS somehow interacts with OraE t o s tabilize the
protein in an active conformation. Moreover, the calcu-
lated isoelectric point of the E component is 9.2, whereas
the S component is 5.1. This result might provide an
explanation for the strong interaction between the S and

E components.
Previous studies have shown that it is n ecessary to include
coenzyme B
12
or B
6
during the re folding of O raE and OraS
in vitro [2]. Both B
12
and B
6
-binding motifs are l ocated at the
C-terminal of OraE and only separated from each other by
about 10 am ino a cid residues. It seems likely that i nclusion
of AdoCbl or PLP during refolding might facilitate the
correct folding of OraE. The d issociation constant, K
d
,for
PLP in
D
-ornithine aminomutase is 224 ± 41 n
M
, indica-
ting that the a poenzyme can bind it with high affinity. It is
not clear whether coenzyme B
12
or B
6
plays a role in pro tein
folding during in vivo translation. To examine this, a m utant

protein, OraSE-K629M, which is unable to bind P LP, was
constructed and produced in E. coli. As the bacterial strain
used to express protein is unable to synthesize cobalamin by
itself, neither AdoCbl nor PLP could be involved in the
mutant protein folding process during in vivo translation.
However, no significant difference in protein solubility
could be found between wild-type a nd mutant protein. This
result indicates that (a) the recombinant protein folding
pathway during in vivo transl ation i n E. coli is different from
the in vitro refolding process, and (b) the a ssociation o f the
S and E s ubunit is importan t for
D
-ornithine aminomutase
to maintain an active conformation in both cases. In
summary, we have successfully constructed, overexpressed,
and purified the recombinant
D
-ornithine aminomutase.
Future work in our group will focus on the determination o f
the quaternary structure of the holoenzyme and th e catalytic
mechanism of this 1,2-rearrangement reaction.
Acknowledgements
This work was supported by grant NSC 91-2320-B032-001 from the
National Science Council, Taiwan, Republic of China (to H P. Chen).
References
1. Somack, R . & Costilow, R.N. ( 1973) Puri fication a nd properties
of a pyridoxal phosphate and coenzyme B12 dependent
D
-ornithine 5,4-aminomutase. Bi ochemistry 12, 2 597–2604.
2. Chen, H.P., Wu, S.H., Lin, Y.L., Chen, C.M. & Tsay, S.S. (2001)

Cloning, sequencing, heterologous expression, purification, and
characterization of adenosylcobalamin-dependent
D
-ornithine
aminomutase from Clostridium sticklandii. J. Biol. Chem. 276,
44744–44750.
3. Baker, J.J. & Stadtm an, T.C. ( 1984) Aminomu tase. In B
12
(Dolphin, D ., ed.), Vol. 2, pp. 203–231. J o hn Wiley & Sons, Inc,
New York.
4. Saito, H. & Miura, K. (1963) Preparation of t ransforming deoxy-
ribonucleic acid b y phenol treatment. Bi ochem. Biophys. A cta 72,
619–626.
5. Higuchi, R. (1990) Recombinant PCR. In PCR protocols. A Guide
to Methods and Application (Innis, M .A., Gelfand, D.H., Sninsky,
J.J. & W hite, T .J., eds), pp. 177–183. Academic Press I nc, S an
Diego, CA, USA.
6. Bradford, M.M. (1976) A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing t he
principle of p rotein-dye binding. Anal. B iochem. 72, 248–254.
7. Tsuda, Y. & F riedmann, H.C. ( 1970) Ornithine metabolism b y
Clostridium sticklandii. Oxidation of ornithine to 2-amino-4-
ketopentanoic acid via 2,4-diaminopentanoic acid; participation
of B12 coenzyme, pyridoxal phosphate, and p yridine nucleotide.
J. Bi ol. Chem. 245, 5914–5926.
8. Schein, C .H. (1989) Production of soluble recombinant pr oteins in
bacteria. Biotechnology ( N.Y.) 7, 1141–1149.
9. Cabilly, S. (1989) Growth at sub-optimal temperatures allows the
production of functional, antigen-binding Fab fragments in
Escherichia coli. Gene 85, 553–557.

10. Totsuka, A. & Fukazawa, C. (1993) Expre ssion and m utation of
soybean beta-amylase in Escherichia coli. Eur. J. Biochem. 214,
787–794.
11. Tang, K.H., Harms, A. & Frey, P.A. (2002) Identification of a
novel pyridoxal 5¢-phosphate binding site i n a denosylcobalamin-
dependent lysin e 5,6-amin omutase f ro m Porphyromonas gingiva-
lis. Biochemistry 41 , 8767–8776.
12. Roymoulik, I., Moon, M., Dunham, W.R., Ballou, D.P. &
Marsh, E.N.G. (2000) Rearrangement of
L
-2-hydroxyglutarate to
L
-threo-3-methylmalate catalyzed by adenosylcob alamin-depen -
dent glutamate m utase. Biochemistry 39 , 10340–10346.
13. Tang, K.H., Casarez, A.D., W u, W. & Frey, P.A. (2003) Kinetic
and biochemical analysis of the mechanism of action of lysine
5,6-aminomutase. Arch. Biochem. B iophys. 418, 4 9–54.
Fig. 5. The plasmid construction map of poraSE. RBS, Ribosome
binding site.
Ó FEBS 2004
D
-Ornithine aminomutase from C. sticklandii (Eur. J. Biochem. 271) 4297